The cytoarchitecture of nerve cells is comprised principally of two structural components: microtubules and neurofilaments (NF). While the function of microtubules in neurite extension and as substrates for transport of particles into the axon is well accepted, no companion function for NF, the intermediate filaments of neurons, is proven. However, we have found that the level of synthesis of the subunit proteins of NF directly correlates with the diameters of mammalian nerve fibers. This has lead us to propose that NF gene expression is a principal determinator of axonal caliber, which itself specifies conduction velocity of the axon. We now propose to use a combination of in vitro and in vivo molecular genetics to test this putative function of NF. Our approach will focus on the NF-H polypeptide, the largest of the three NF subunits. While the other subunits (NF-L and NF-M) are more abundant, the long carboxy-terminal tail domain OF NF-H has long been speculated to cross link adjacent NFs and perhaps to determine the lateral spacing of NFs. Starting from the cloned mouse NF-H gene, we shall use DNA transfection to determine the assembly properties of NF-H when expressed alone or in combination with NF-L and NF-M. Since in axons NF-H is heavily phosphorylated and its carboxy-terminal domain carries >50 repeats of a putative phosphorylation site, we will also perform similar experiments with mutant NF-H genes deleted in this and other domains. Finally, the in vivo properties of such mutant NF-H polypeptides will be assessed by producing transgenic mice whose genomes contain genes specifically engineered to increase or decrease NF-H expression.